Please see this article that describes the AWM2 Envelopes: Although written for the Motif XS/XF and MOXF, everything still applies in the current AWM2 engine. If you still have questions, post back here.

Link: Learning About Envelopes

Still can't get close to what I want.

Here's the desired outcome, in a visual form:

Can you suggest how this could be created, on an underlying (by way of example) AWM sound that's of a long and sustained volume, such that it sounds like this kind of pluck, regardless of how long the key is held by the player?

The illustration you gave is a generic analog exponential curve envelope shape, it doesn't fully explain what you're trying to achieve.

Are you trying to use the AWM2 EG as a regular ADSR, or are you trying to achieve the exponential curve shown?

The AWM2 envelope has separately settable times and levels for each stage: Initial (level) or for the filter EG, Hold (time+level), Attack (time+level), Decay1 (time+level), Decay2 (time+level which is also the Sustain level), then Release (time only for Amp EG, time+level for Filter EG). Another difference between the Amp and Filter EG are the levels are bidirectional on the latter, meaning they can be set to positive or negative values.

To emulate a regular ADSR, set the Levels and Times as shown below:

LEVELS: Initial or Hold = 0, Attack = max(+127), Decay1 and Decay2 both the SAME level, corresponding to your ADSR's Sustain level, for a pluck set both to 0, Release = 0 (filter EG only)

TIMES: Hold (Filter EG only) = 0, Attack = your desired attack, 0 or very low for a pluck, Decay1 = your desired decay time, Decay2 = 0, Release as desired (same as Decay1, or shorter or longer as needed).

The curves are linear in AWM2, so it may not sound as "plucky" by default as an analog ADSR. You can get a bit closer by using the 2 decay stages, with the 2nd decay a little slower (higher value) than the 1st, and the Decay1 level set to a low value, and Decay2 = 0. Playing with mine, I get a pretty good snap and still have an audible decay like an analog envelope (think Popcorn) using:

Times: Attack = 9, Decay1 = 35, Decay2 = 40, Release = 40
Levels: Initial = 0, Attack = 127, Decay1 = 12, Decay2 = 0

You could search ( [CATEGORY SEARCH] ) for a factory Preset with the name "Pluck" in it - there are a few - and look at those elements' Amplitude Envelopes.

You could download scope software and look at the waveform over time to see how Amp EG changes come out the final output. As you know, the original sample has its own envelope inside the recording. So depending on the sample, the resulting envelope will differ.

I used the Alto Sax preset - which holds the note "forever" if you press a note. The sample doesn't die out on its own. That was part of my criteria. Then I applied a couple envelopes. The first one I did by ear - which made a nice plucky sax. The second one, I fooled around with attack and decay1 until I got the attack more rounded out like your example image. I don't like this one as much - but it shows you have controls.

This is windows software and the name is in the picture.

Or you could just use your ear until it sounds right. This is what I always do. Typically I'm never quite satisfied with the "perfection" of my envelopes. But they're more than good enough for the audience - so I move on.

Or you could just use your ear until it sounds right. This is what I always do. Typically I'm never quite satisfied with the "perfection" of my envelopes. But they're more than good enough for the audience - so I move on.

I have been unable to get closer.

Can you give me some clues as to what your settings were to get these curves?

For 2nd response curve:

Started at level 0. Typical sharp attack - attack only brings up to part of the way before that more gentle slope going up and to the right at about a 45 degree in the picture. This "45 degree" is decay1. Decay 2 covers the next section and end of decay2 is back at level 0. I'm not sure I saved any of this. But it took some trial and error using the software to visualize the waveform and adjusting primarily attack vs delay1.

Did you look at any of the "pluck" factory Performances to see if the envelope response matched what you're after?

If so - those can be used as guides.

Yes, long before I posted this question, and from the first day of getting into it... I looked at the presets.

Most of them use samples that have a lot of the "attack and falloff" in the sample.

Hence the question's context, of using a sample that's a solid, non-decaying sound.

Still not getting a desirable sound.

What do you mean by decay 1 being part of the attack?

A plucked envelope would require Decay 2 Level = 0

Decay2 Level = 0 in your MODX amplitude envelope generator describes a hammered, struck or plucked envelope.
Hammered (piano, is an example), struck (vibraphone or a drum, is an example) or plucked (guitar or harp is an example) are the percussive family of musical sounds. The musical vibrations are put in motion by this triggered event. And no matter what you do after having hammered, struck or plucked the instrument, the sound will eventually die out. Even using the “sustain” pedal on a piano does not prevent the vibration from eventually stopping. A sustain pedal (so-called) only delays the inevitable fate of all musical events in the percussive family... they return to silence on their own. It is the hidden forces of gravity and air friction that end the vibration.

The other major family of musical instruments are described as “self-oscillating” instruments... these initiate the musical vibration by bowing or blowing. These have envelopes where Decay2 Level is not 0. The note event ends only when the musician stops putting pressure on the instrument. These self-oscillating instruments can maintain the musical vibration, ‘forever’... they can defy the inevitable return to 0

Bowed (a violin, viola, cello are examples), blown (a clarinet, flute, oboe, trumpet are examples - these make up the reed family - lip reeds, single reeds, double reeds, jet reeds.
Bowed instruments can continue sounding by artfully changing the direction of the bow. They are not limited to a single trigger event (as in percussive envelopes) - they can maintain pressure by continuing to move the bow across the string. Blown instruments can continue sounding by artfully circular breathing. This is where the player uses the lungs exhaling to start musical vibration, they can maintain that outward pressure by collapsing the cheeks while taking air in via the nose (refilling their lungs).

In the self-oscillating instruments the player decides when Decay2 Level reaches 0 (not gravity, not air friction). Whenever they stop applying pressure - which translates to KeyOff on your synthesizer, the AEG immediately advances to the Release Time setting (how the sound will disappear after KeyOff - how it returns to silence).

Try to understand the stages of your synths envelope in terms of these two fundamental families.
Attack describes the start of vibration. Immediate or very near immediate, best describes a percussive attack. It spikes in amplitude at the attack, and often dies to a secondary stage of loudness. Decay1 Level

Picture a bass drum being struck a mallet... the trigger event spikes as the contact is made... it reaches its loudest almost immediately. The percussive attack is typical heard as a ‘click’ or ‘snap’ at about 1.2kHz through 2kHz. Then the actual size of the bass drum comes into play as the ‘boom’ -- this is described by the envelope moving into the longer low frequency response of the bass drum.. while loud, it is not as loud as that initial attack. That initial attack is called a “transient peak”. This second envelope stage called the Initial Decay can be described in the MODX using Decay1 Level

All hammered, struck and plucked instruments have this “transient peak” located at the front end of their amplitude envelope... and a varying length stay at Decay1 Level. A church bell that rung is loudest at the Attack... it then settles in to its long sustained ring (Decay1 Level) slowly losing energy to gravity and air friction which act to stop it.

Contrast this to the beginning of bowed and blown instruments. While percussion instruments are like vocal “plosives”, bowed and blown instruments can start gradually like words that start with the letters “s”, “f”, etc... ”smooth”, ”freedom”, the sound gradually rises up from silence. Unlike the percussive word beginnings like “petunia”, or “boom” which explode with an initial transient attack before the milder sustain portion. The bowed or blown instrument can maintain the exact same volume throughout, they can even adjust that volume during the sustained portion... which can be truly sustained at the same volume if desired... all the way through Note-Off.

If your goal is to create an amplitude envelope that dies out in spite of the key being held... then Decay2 Level must be 0.
And it must arrive their by first traversing the other stages that precede it.

But what about those other settings?
What you need to get a sense of is the TIME parameters... if your TIME parameters are too long, your constructed amplitude envelope may take too long to develop. With TIME parameters the higher the number the longer it takes to traverse this portion of the envelope. Even if you set Decay2 Level = 0 (which we are saying defines “percussive behavior) not all percussive sounds are quick.

Think of a large bell... it is struck once and can ring for more than a minute on a single strike. This is because the TIME settings prior to Decay2 Level are set very high...allowing a long slow AEG.

TIME is how long it takes to travel from one point to the next in the envelope. The higher the number the longer it takes to travel...also how far it travels, or has to travel, also influences your result.
LEVEL in this case, is how loud it is at that time.

Suggestion: Avoid extremes - change settings gradually - they values are NOT linear. Repeat - they are NOT linear! These are constructed specifically to be musically relevant— and are weighted toward where they would be most useful.

Call up the MODX Preset Performance: “Bass Morpher” — made from sampled (Prophet V sawtooth waves)
In the screenshot below I have selected Part 1
Dropped into [EDIT]
Muted Element 2, 3 and 4 so we can concentrate on just Element 1 (It is important when working on envelopes to isolate the item you are working on - makes it easier to hear your changes.

Navigate to “Elem1” > “Amplitude” > “Amp EG”

As you move the cursor to a Time or Level parameter notice the white pixel in the graphic will indicate where in the AEG you are.
Amplitude Level is the y-axis (up/down) — Time is the x-axis (left-to-right).
I highlighted Decay1 Time = 64

As you slowly increase and/or decrease this parameter value notice the change in behavior of the envelope... re-strike notes to restart the AEG
You can lengthen the Time, the white pixel moves to the right along the x-axis, the sound takes longer and longer to return to 0 but you cannot sustain it forever, because eventually the LEVEL parameters dictate that it will disappear (reach the bottom or floor of the graphic.

A sound will sustain ‘forever’ if Decay2 Level is set to a significant value and you are holding the KEY or the Sustain Pedal (which takes the place of holding the KEY). ‘Forever’ in quotes because in music that is a fictional quality... we know it as applying pressure to the instrument. If you stop applying pressure to a musical instrument, the sound moves immediately to Release Time... Release Level is not a necessary parameter because it is “musical” silence.

What do you mean by decay 1 being part of the attack?

The easiest way to create a plucked sound is to make the attack level the "peak" of the audio amplitude. And then just fall off from there. That's how the 1st upper-left green graph was constructed.

Your original waveform picture (upper-right in my picture) slowly "rolls off" after a sharp attack. So I approximated this by having the attack level as NOT the maximum level of the envelope. I went up to around 80% then from 80% to 100%, I used decay1. The end of decay1 is the peak in the lower green envelope picture. Then decay2 drops off, down to 0, from there.

I personally don't like the sound of this better than the easier approach - because it's not as snappy. But it matches your picture better.

Ok. Interesting. I'm still (even after enormous readings, HUGE amounts of trial and error) unable to understand what decay1 and decay2 actually are.

If asked, by someone, to explain these, I couldn't.

That's weird, in and of itself. I'm not normally this dense.

For example, I'm still unable to get the pluck to behave the exact same, regardless of how long I press the key.

I don't want my key press duration to impact the pluck, only the velocity of the key press.

Is this possible?

Is it possible with the kind of curve of pluck volume over time I've drawn?

Velocity variation in the envelope is a different thing. Myself, I've zero'd out all velocity based variation in the envelope. Although, it's possible to affect the envelope with velocity as an input - and these parameters are built straight into the AEG menu. I wouldn't advise graduation to velocity variation until you have non-velocity variation straight first.

Start with "Init Normal (AWM2)". This initialization is a piano sound - but the rest of the settings are very vanilla. There is no velocity sensitivity in the response of the single element that sounds in this initialization Performance. I can't remember if there's velocity variation in the AEG or if it needs zeroing out or not - but these settings are easy to recognize and you can zero these out if they are not already.

The main advantage to using this initialization Performance is that there will be no "Mod/Control" -> "Control Assign" source/destination pairs that could foul up your settings. Often when you inherit someone else's programming - you have to deal with what initially seems like a mystery novel from the "Mod/Control" -> "Control Assign" menu. Using the initialized Performance won't have these - so a more deterministic starting point for learning the ropes.

Note that the response waveforms I have given will be the response from holding down a key "forever". The AEG is what creates the pluck - not the duration of a key press.

AEG is detailed on page 33 of synthesizer_en_pm_c0.pdf

It shows that when you press a key that will trigger an element (called "key on" event), the said element will start sounding at the level set by "Level: Initial".

Assuming you keep holding down a key, the element will take "Time: Attack" in order to change the amplitude to the next inflection point "Level: Attack". One point to make - the graph in the documentation shows the initial level (at "I" ) is less/lower than the attack level (at "J" ). But it doesn't have to - the initial level can be higher than the attack level. Keep this in mind not to take the drawing too literally in terms of conventions for HOW the AEG is drawn. It's merely one typical representation of AEG - but you can set these differently and rise and fall differently than the picture.

Next, after the attack level is reached - the element will take "Time: Decay 1" to change the amplitude to the next inflection point "Level: Decay 1".

Next, after the Decay 1 level is reached - the element will take "Time: Decay 2" to change the amplitude to the next inflection point "Level: Decay 2".

The description labels "Level: Decay 2" as the "sustain level". What this means is that if you keep holding down a key - then after decay 2 is reached, the amplitude will remain at this level for as long as you hold down the pressed key (note on). For samples that are looped (like organ samples, the alto sax, etc) - you will hear the sample sound at "Level: Decay 2" and this will remain constant.

If you make the Decay 2 level 0 - then this is the absolute requirement, when using looped samples, to get the "pluck" effect. The rest before this is gravy. But if Decay 2 is 0 - then, eventually, if you hold down a key forever - the element's amplitude will diminish to 0 as a consequence of the AEG. It sustains at 0 volume. The only responsibility you would have before Decay 2 Level = 0 is to make sure the time values are not too long.

Finally - when you let go of a key - it will start at whatever level the AEG is at when you let go (maybe 0, maybe not) and take "Time: Release" to change the amplitude to the last inflection point "Level: Release". Which, in AWM2, is fixed at 0. This is the one part of the graph in the docs that is set in stone. AWM2 AEG always ends in 0 amplitude. So everything will diminish eventually.

If you let go of a key - then the release will happen starting at whatever level the AEG is at during the moment of key release. So if you have a very slow attack, and the attack level is 127 (maximum) then it may take some time to reach 127. On the way up to 127 - the AEG will slowly rise and perhaps half way up a slow attack at around 64 - you let go of the key. The release will start at the level of 64 and then diminish to down to 0 after "Time: Release".

But for what you're after - attack will be fast. And relatively high level. Decay1 can keep going up at a slower slope or not. You can make decay1 a level of 0 if you want. Or you can use Decay 1's level to "shape" the tail. Decay 2's level must be 0. Release's level is always 0. Set the release time to 0 or something small so that if you do play staccato - the release will be short.

For example, I'm still unable to get the pluck to behave the exact same, regardless of how long I press the key.

I don't want my key press duration to impact the pluck, only the velocity of the key press.

Attack, Decay1 and Decay2 take effect while the note is being played (key down or sustain pedal held... in MIDI terms this is called "note on").

When the key is released (and sustain pedal is also released), the note transitions to the "note off" state. At this point the Attack, Decay1 and Decay2 TIMES no longer apply to the envelope. Instead, the RELEASE time kicks in immediately and the envelope slope changes to reflect the release time parameter.

If you're using Decay1 and Decay2 together to shape the decay of your note, you won't get the exact same effect with the single Release time if your note duration is too short. You'll have to experiment with the Release time (start with it being set to the sum of Decay1+Decay2 times, then adjust by ear), to make the note decay the way you want when playing short duration notes.

Also, if you're using the Time/Key parameter to vary the speed of the envelope based on the note being played, you'll want to adjust the Release Adj parameter so the Release time adjustment across the keyboard matches the Decay time adjustment. Set this to 127 initially, and tweak by ear.

@Kevin...

yes... I'd sort of come to a very unsatisfactory "solution"...

if the Attack time is 0, Decay1 time is exactly half of the spread available (63) its value is 0, and the Decay2 time is 0 and its value also 0, and the Release is exactly 63, then this is the only time that, regardless of how long I press the key down, the falloff is the same.

Any other setting, and a time of note duration that's less than the completion of Attack, Decay1 and Decay2 results in the Release initiating from whatever volume state the envelope had reached at the point of note release.

Not what I wanted.

If you wanted to completely ignore the note off event, you could look at using a drum part instead, though using that for a pitched instrument would be very annoying.